It’s time to move away from pigeonholing data as good or bad

Charlie Spargo's picture
by Charlie Spargo

Barley Laing, UK Managing Director at Melissa, looks at why data shouldn’t be perceived in simple ‘good’ and ‘bad’ terms, and how brands can identify and unlock the potential value of ‘bad data’ to access the valuable insight it contains.

What is ‘good data’? What is ‘bad data’? It’s not as easy to define as you might think. Just because data is old or missing an element doesn’t necessarily mean it’s bad and should be disregarded or deleted; nor is freshly collected data always good, particularly if it has not been properly verified at the point of capture.

In today’s big data world, it’s difficult to classify data in such a binary way, especially when the driving force behind delivering improved marketing and business success is derived from effectively analysing and cleverly using customer data. 

The definition of bad data is up to each individual organisation. But if a customer record is deemed bad because it’s missing a phone number, the address is wrong, or an email is blocked - should that entire record be purged? The short answer is no! 

Data that is simply incorrect, such as customer name, address, email or telephone number, is an issue that can be easily fixed.

Clean customer data

Without regular intervention, customer data degrades at 2% each month, and 25% over the course of a year - a major issue for brands. Additionally, in an age when customers are increasingly providing contact data via their mobile devices, mistyping contact details on a small screen has also become a problem.

In fact, approximately 20% of addresses entered online contain errors such as spelling mistakes, wrong house numbers, and inaccurate postcodes.

Clean data can be delivered via industry-leading data cleansing, standardisation and verification services that provide quality data in batch form and data entry verification in real time. A solution that’s scalable, can integrate with your CRM platform and can cleanse, correct and format UK as well as global name, address, email and phone numbers is ideal.

Cleaning data this way is an excellent first step to unlocking the insight harnessed within bad data and helping to make it good. 

Address autocomplete for accuracy

To prevent collecting bad or inaccurate customer address data, leverage an address autocomplete service. This type of resource automatically recommends the correct version of the address, in real time, as the customer completes an online contact form, promoting the selection of the one that’s not only accurate but easily recognised.

It solves the issue of mistakes caused by ‘fat finger syndrome’. Such a service also reduces the number of keystrokes required when typing an address by up to 70%, accelerating the checkout process and lowering shopping cart abandonment. 

ID verification prevents bad data and fraud

To avoid bad data at the customer onboarding stage, it’s important to take customer checks and verification to another level, not only for data accuracy and to deliver good data, but as protection against fraud. This is key as escalation in the number of data breaches has led to an increased number of criminals posing as legitimate consumers, albeit with stolen or falsified identities.

Delivering effective customer verification all comes down to data. This requires sourcing a global dataset of billions of records containing data from trusted reference sources such as credit agency, government agency, utility company and international watchlist data.

To ensure customers are who they say they are, brands must carry out cross-checks of captured customer insight against this data; matching a particular name to a physical address, telephone or email. Furthermore, such a dataset can also be used to verify the end user’s age to ensure they are legally entitled to the product or service offered. 

When sourcing data for ID verification, it’s important not only to verify but also to enrich and improve the customer data to deliver a 360-degree single customer view (SCV). This SCV can also aid future marketing and sales efforts. Since consumers expect a seamless onboarding experience in this increasing digital age, it’s vital that the technology powering the ID verification process delivers this insight in real-time.

AI semantic technology

Artificial intelligence (AI), in the form of machine reasoning semantic technology, is another option to consider in delivering good data. Semantic technology, or semtech, associates words with meanings and recognises the relationships between them.

It works by delivering powerful real-time connections between customer records - combining the missing pieces of customer data to improve data quality, confirm identity, and make an informed decision about whether to present a particular product or service to a customer.

Machine reasoning does this by filling any gaps in information left by the customer during the onboarding process or via other communications. AI not only improves data quality, it delivers the information that empowers organisations to make informed decisions around the products and services it offers to customers. 

With definitions of good and bad data so subjective, the only path forward is to ensure all customer data is effectively cleaned and verified. Today, as data services and technology evolve, it’s possible to clean, verify and fill in the gaps of what is perceived as both bad and good data, then generate actionable insight from it in real time.

With brands handling increasing amounts of customer data, it’s crucial to take these steps to generate as much value as possible from the data to ensure you stay ahead of the competition.